Treatment response in COPD – does FEV1 say it all? An analysis of the CRYSTAL trial

Rationale: Improvements in FEV1 correlate with patient-related outcomes (PROs) at a population level (Donohue et al. Pulm Pharm Ther. 2018). However, the association of clinically relevant changes in PROs with changes in FEV1 in individual patients have seldom been investigated. Methods: CRYSTAL, a 12-week, prospective, randomised, pragmatic trial, evaluated the effect of indacaterol/glycopyrronium (IND/GLY) or GLY after direct switch from previous treatments in moderate COPD patients. Lung function (FEV1), COPD assessment test (CAT), clinical COPD questionnaire (CCQ) and Transitional Dyspnoea Index (TDI) were evaluated after 12 weeks vs baseline. Results: Of 4324 patients 2774 (64.2%) were classified as responders by TDI, CAT or CCQ with 1365, 826 and 583 patients responding according to 1, 2 or all 3 PROs respectively. In comparison, 3235 patients (74.8%) were considered responders when combining TDI, CAT, CCQ and FEV1 and 1555 patients (36%) responded to PROs only (≥1) but not to FEV1. Out of 1680 responders to FEV1, 1219 (72.6%) patients demonstrated a combined response (FEV1 and ≥ 1 PRO), and 461 (27.4%) showed only a response in FEV1. There was little overlap between responders to TDI, CAT, CCQ and FEV1 (figure). Conclusions: Improvements in lung function do not always translate into improvements in PROs in individual patients. PROs are not interchangeable and hence the definition of treatment success depends on the PRO used.